"You rapacious information and identity thieves better get ready for me to go full McConaughey on you. Also, you suck."
That was tech journalist Kara Swisher's response to discovering that Grammarly had been selling AI-generated writing advice under her name - without asking her first.
On March 11, Grammarly disabled its "Expert Review" feature after a wave of anger from the writers it impersonated and a class action lawsuit from investigative journalist Julia Angwin. The feature, available to Pro subscribers at $12/month, generated writing feedback framed as advice from specific real people: Stephen King, Carl Sagan, Kara Swisher, Platformer's Casey Newton, copy editor Benjamin Dreyer, and others.
The problem was simple: none of them agreed to it.
How the Feature Actually Worked
Expert Review, launched in summer 2025, let users pick a named expert and receive AI-generated critiques modeled on that person's writing style and editorial perspective. The feedback was not written by the experts. It was not reviewed by them. It was not endorsed by them. In many cases, they did not even know it existed.
Benjamin Dreyer tested the feature himself with lorem ipsum placeholder text. It still generated "tips from writers including the venerable novelist Stephen King." The advice was generic regardless of whose name was attached to it, which made the use of real identities feel even more gratuitous.
Casey Newton summed up the absurdity: "I just assumed that someone would tell me when it happened" - the moment an AI version of himself started giving writing advice to strangers.
Grammarly's original policy required experts to opt out manually. You had to discover the feature existed, then take action to remove yourself from it. That approach did not survive contact with the writers it affected.
The Lawsuit and the Retreat
Julia Angwin filed a class action lawsuit against Superhuman, Grammarly's parent company, alleging unauthorized use of writers' names and likenesses in violation of privacy and publicity rights. "I have worked for decades honing my skills as a writer and editor, and I am distressed to discover that a tech company is selling an imposter version of my hard-earned expertise," Angwin said.
Superhuman CEO Shishir Mehrotra acknowledged the criticism and announced the feature's removal. "We hear the feedback and recognize we fell short on this," he said, adding that the company would "reimagine the feature to make it more useful for users, while giving experts real control over how they want to be represented - or not represented at all."
The phrasing is worth reading twice. Mehrotra is not saying the concept was wrong. He is saying the execution was wrong. Superhuman still believes AI can simulate expert-style feedback - they just need to get consent first.
That distinction matters for every AI company building features around real people's identities. The technology to mimic someone's expertise is here. The legal and ethical framework for doing it without permission is not.